PurposeThe purpose of this study is to examine the impact of ownership structure such as family, government, institutional, foreign and public on dividend payouts as a representative of dividend policy of nonfinancial firms in Bangladesh.Design/methodology/approachThis study employs a dynamic panel data model, namely, differenced generalized method of moments (GMM), which follows a two-step process. The study uses annual data of a sample of 159 nonfinancial firms of Dhaka Stock Exchange for the period 2008–2017, which constitutes a panel data of 1,590 firm-year observations.FindingsThis study’s findings reveal that family and public ownerships have a significant and positive effect on dividend payouts, while government and institutional ownerships have a significant but negative effect. This study additionally incorporates some very important controlled variables and finds that except for size, all the selected controlled variables, i.e. lagged-one of dividend payout, returns on assets, debts to assets, price-earnings (PE) ratio, age and financial crisis have a significant effect on the dividend payouts. However, the findings support several dividend-related theories or hypotheses, i.e. agency cost theory, dividend stability theory and reputation hypothesis.Research limitations/implicationsThis study could consider some other aspects of corporate governance, as well as other emerging markets and financial institutions to perceive whether the results differ. Also, investigation could be carried out on conventional and Islamic firms individually to observe if the findings are different. However, the researchers are suggested to incorporate these issues in their future studies.Practical implicationsThis study offers an important insight into the relationship dynamics between dividend payouts and ownership structure in the context of an emerging market like Bangladesh. Moreover, it enhances the understanding of the ties of dividend payouts with the firm-specific factors as well as the financial crisis. The findings of the present study have also important implications for managers, policymakers and researchers, who are in quest of directions on the dividend policy of publicly listed nonfinancial firms.Originality/valueMost of the previous studies consider one or two types of ownership to examine the impacts on dividend payouts, while this study uses five types of ownership accompanied by a different data set. Moreover, to the authors’ knowledge, no study in Bangladesh has yet addressed this issue in such a comprehensive manner as theirs.